Tech Deciphered podcast

83 – Smoke and Mirrors: How the Communications Machine Took Over Tech

19.11.2026
0:00
50:54
15 Sekunden vorwärts
15 Sekunden vorwärts

The mythology in Silicon Valley is well known: stories that don’t quite match reality. Zero to Hero and Hero to Zero narratives, and everything in-between. This episode will focus on the communications machines of large and small Technology companies and its Investors, which have taken over the narrative, sometimes with disastrous consequences.

Navigation:

  • Intro
  • The Factory
  • The Performance
  • The Megaphone
  • Our Take & Conclusion

Our co-hosts:

Our show: Tech DECIPHERED brings you the Entrepreneur and Investor views on Big Tech, VC and Start-up news, opinion pieces and research. We decipher their meaning, and add inside knowledge and context. Being nerds, we also discuss the latest gadgets and pop culture news

Subscribe To Our Podcast


Nuno Gonçalves Pedro

Introduction

Welcome to Episode 83 of Tech Deciphered. This episode will be about communications. We’ve called it “Smoke and Mirrors: How the Communications Machine Took Over Tech.” We will go through several examples on how the mythology of Silicon Valley has been built over the last few decades—from basic accelerators and how those companies have been able to create the notion of controlled access, to all the way through the tech companies, the big tech companies, and their communication machines, which are not always the most truthful. As always, no BS, straight to the point. Bertrand?

Bertrand Schmitt

The Factory

It’s been around for a long time, so in a way there is nothing new. What might be new is the scale, the experience in all this marketing that different actors have, how this has been structured for a while. You could argue one of the most talented communicators was Steve Jobs, who started Apple, managed its comeback. He was considered to have one of the most potent weapons: the “reality distortion field”—his ability to make you believe in nearly anything. Usually it was, for sure, very surprising, impressive, but not always. But one of the most talented, gifted practitioners in that space.

Nuno Gonçalves Pedro

Yeah, even Bill Gates, right? The whole word “vaporware” in some ways, I think, was—I’m not sure historically this is totally true, but it was, in my opinion, seemingly attached to Microsoft on how it announced certain things that really didn’t quite work out the way it was supposed to. That was back in the good old days of Bill Gates running the company.

Bertrand Schmitt

You’re right. At some point, it was called “slideware” from Microsoft.

Nuno Gonçalves Pedro

Which is appropriate, given Microsoft’s ownership of PowerPoint, yes.

Bertrand Schmitt

Exactly. Yes, nothing new under the sun, but I guess it’s a good time to talk about this and explain a bit some of the recipes.

Nuno Gonçalves Pedro

Maybe let’s start with the beginning of the problem, so to speak. If we go through stages of funding in some ways, and the beginning is the factories—the accelerators, what we would call the “demo day theater.” There is, in some ways, an industrialization of startup creation happening, and the accelerators—or certainly certain accelerators, Y Combinator being probably one of the strongest examples of that—have become a signaling machine, so to speak. Not really true to me that all of these accelerators are schools.

In some ways it feels like the education they’re providing is on how to make yourself a hot company out there, and how to really propel yourself into fundraising—those two key things: communications plus fundraising as part of that offering. Y Combinator has grown quite a lot. We were just checking how many batches they have. They have four batches a year: they have spring, summer, fall, and winter. The spring, I think, is called “X,” just not to be confused with summer, which is “S.” Summer ’26 would be S26—that’s the batch—but then spring ’26 would, I guess, be X26, not to be confused.

They have a ton of companies. The current cohorts are each roughly half the size of the old mega batches, but to be honest, when Y Combinator started, they didn’t have mega batches; the mega batches came later as well. But summer 2026, if we have the numbers absolutely right, had 235 companies in that batch alone. It feels like a batch machine. For YC, it’s great. Because they can cherry-pick. They’ll get into all the companies, but then at some point they can cherry-pick which ones are really great and not. They seemingly still have a little bit this index way of attacking the industry and really creating some of the biggest successes at the same time as the others that fail through the wayside.

Bertrand Schmitt

They still have a pretty high acceptance rate—even when you hold the size of the batches comparable, it’s a 1% acceptance rate. They are still pretty careful. It’s pretty impressive, I would say, to see that. After that, there is always this hype machine in terms of: do you manage to get in before the demo day, or just on the demo day itself. It’s always quite impressive. I would say some will say that the really big successes have been a bit more few and far between recently. I mean, compared to the origins of Y Combinator—the first few years—you might argue there might have been less, much bigger successes than before. But obviously they have a capacity to still attract quite a lot of pretty successful startups, or startups that are going to be pretty successful.

Nuno Gonçalves Pedro

I think YC is probably the most extreme example. It’s based in Silicon Valley. It feels like the Silicon Valley accelerator. Obviously there are others; 500 Startups is one of them.

Bertrand Schmitt

They do still exist, 500 Startups?

Nuno Gonçalves Pedro

They do. I’m not sure—let me put this in another way. I feel they’ve gone very geographical. They’ve decentralized for sure, and I’m not really sure how much of it is batch playing all the way through fund management. At the end of the day, they seem to have different interests. In some ways it’s a little bit like a franchiser-franchisee relationship, which Y Combinator is not. We have others—Techstars, Dreamit, et cetera. But they’re all working on the same thing: on distilling everything to a three-minute demo, on trying to get the companies and startups that are in the batches the highest chance of fundraising out of it as possible.

Again, a lot of it is around communications—around a specific kind of communication. I was having a meeting the other day with a founder. He was very pissed off at me, I think. I wasn’t being facetious or aggressive, but he was pitching to me as a VC, and he came out of Y Combinator, I think, a couple of months before. I was like, “Hey dude, you’re giving me the demo day pitch. Not interested.” I’m not interested in the three-to-five-minute high-level thing.

I’m like, “You were kind to share your deck. I’ve looked at your deck. I have specific questions. Can we go into that?” He kept going back to his message house. That’s how powerful the embedding—it’s like someone at Y Combinator or something, I don’t know, told him, “You need to stick to your message.” I’m like, “Well, I’m a VC, and the first thing I’m going to do is destroy your message house, because I want to get to the risks that your company has right now, and how are those risks measurable in terms of our analysis of your company, whether this makes your company interesting or not.”

In the end, obviously—as you can imagine—we passed not so much because of the grand posturing of the thing, but because we didn’t believe the company was really at the right stage for us to come into. But that was very visible. He kept going back to his message house. I’m like, “Dude, I understand what you’re doing. I’ve been to a lot of demo days—Y Combinator and otherwise. I understand, but I have questions and we have 30 minutes. Can I go through my questions?” “No.”

He wanted to go to his message house. In the end, I ended up putting some questions at the end after 15 minutes of him grandstanding with several layers of his message house. I was like, “Dude, not for us, I think. Not for us.” sometimes over-communicating—sticking too much to what you were told—doesn’t work. There’s something that you learn in communications and in negotiation very early on, which is the notion of releasing agenda. If the other person pushes you, probably the best option is to release your agenda, not stick to it. If you stick to it, then I’m like, “I’m not paying attention anymore.”

Bertrand Schmitt

My guess would be most of the learnings of Y Combinator have definitely some value, but at the end of the day, you need to adjust to your audience. You need to adjust to the needs of your audience. If you work like this with investors, then you will work like this with customers. At some point, that’s not going to work. Maybe there was not a lot of substance in that specific situation.

Nuno Gonçalves Pedro

There’s a couple of other accelerators we haven’t really mentioned. Speedrun from A16Z. Big batches now—they’ve moved as well from what used to be mostly gaming at the beginning and consumer to something that’s a little bit more broad. A little bit—no, a lot more broad. We’ll see how that works out. Early on, I remember going to demo days, and they were very well organized, very well done. They seemed to be going for the kind of company that already has some traction, which is, by the way, what 500 Startups used to do. That was one of the key distinctiveness versus YC, which seems to be more like earlier startups.

Then there’s more specialized accelerators out there that have done very well. HF0 is one of them. Relatively hot serial entrepreneurs and all the solutions around it. There’s a few others that I’m probably not remembering exactly the name right now. This is on top of all the global, regionalized, domestic incubators and accelerators that exist out there, which are many.

But I feel we’re moving too much towards, less towards the beginning of this, which was around understanding markets, product, how to get you to product-market fit, or some semblance of product-market fit before demo day, in particular if you’ve already launched. More around the craft of how do you build products, and you achieve product-market fit. We’re moving more and more from that to communication—to communication in order to raise. Once there’s money, then you’ll be able to get through that. I’m like, “Well, I’m not sure that’s the right solution for the industry.”

Bertrand Schmitt

Yeah, you’re right. It used to be more about helping you find the right product-market fit, and hopefully you achieve that before demo day. The change you describe is interesting. I don’t know if it’s truly surprising, but it’s interesting. Another thing I noted with YC is that there is always a theme. If you want to check the pulse of the industry, you check what type of startups do you see in each batch. Is it all about mobile? Is it all about SaaS? Is it all about AI? Is it all about robotics? Because usually it provides some good perspective about where the market is at, at least in its early stages.

Nuno Gonçalves Pedro

Some of this works. One of the better-known studies out there, I think—Hallen, Bingham, and Cohen—that compared accelerated startups to ones that didn’t get in or that almost got in, and found that accepted startups were generally more likely to be alive or acquired down the road, had more employees, and were more likely to have raised over 1 million in, let’s call it, VC, micro-VC money. There’s clearly something attached to this. It’s a signaling piece that is significant. The startup that comes out of Y Combinator has a badge. It’s a little bit like in the good old days of career-making.

If you had worked for McKinsey, that’s a badge. That matters. Branding matters. In some cases it will facilitate your next one, maybe two rounds. But at some point the rubber meets the road, which is: what is actually your business? Are you about to hit product-market fit? Is there something that fundamentally shows that because you did Y Combinator, you are better than others? I am not fundamentally sure about that. It’s a great signaling mechanism. It facilitates a lot for many.

There’s a lot of firms right now and funds that are focused a lot on Y Combinator. But question mark whether that will continue holding true, if what we saw in the past of great performance by Y Combinator batch companies will hold true in the future, if this new wave of four batches a year, a lot of quantity of startups, et cetera, will actually work out. I suspect not. Because just looking at quantities, if Y Combinator were to process around 800-1,000 companies a year, I don’t think there’s space for 800-1,000 companies that will be incredibly performing on a yearly basis in the industry. That alone would be a problem.

Bertrand Schmitt

Yeah, another thing I notice is that there is not a lot of very famous AI-focused companies coming out of Y Combinator. You have, at best, maybe Replit. But even them, they switched the business model. That’s not where it started. Funny enough, they put actually OpenAI front and center now on Y Combinator homepage. But if you read the fine print, it’s because Sam Altman’s previous startup was part of YC inaugural batch in 2005. Then he founded OpenAI—that’s when he was already leading Y Combinator. That was not your typical Y Combinator startup. He left Y Combinator in order to focus on OpenAI.

For me, I find it interesting because to put that as the first company you showcase. I can imagine you want to benefit from the brand of OpenAI, but at the same time, it’s a fake in terms of what it is going to bring and what it means for investors. Because if you participated to Sam’s first startup, that might not have gone as you wanted. OpenAI was not part of demo day. You could argue that raised some questions about where this is going. Is an accelerator the right way to push things forward in the age of AI, especially a generalist accelerator?

Nuno Gonçalves Pedro

Recently we’ve had a lot of companies that have leapfrogged accelerators and raised a ton of money. Frontier Labs are a great example of that. Most of them didn’t go through any acceleration. They just raised a ton of money from VC straight up, from strategics in many cases as well. But again, that connects back to the world we’re in—the world of everything that shines AI with uniqueness, distinctiveness on talent, seems to be able to get funded. As we’ve discussed in previous episodes, we’re not sure that will continue for many years to come, but certainly this is where we are today.

Bertrand Schmitt

If you look at Y Combinator again, homepage, and how they communicate about their best success stories they keep showing you pictures of very young people, 22-year-olds or so, and no offense, it’s great to start a company at that age and deliver and execute. The problem is that we might be in a different world. As you say, the people who built some of these labs, they were not 22-year-olds. They were typically 30, 35, 40, already very well established in the industry, having done their marks. They didn’t need the Y Combinator imprimatur to be able to raise money. That’s actually a pretty big difference that the industry might be changing.

Nuno Gonçalves Pedro

The Performance

Maybe moving to the performance. The next stage—when you have founders that are influencers, that are building in public, or building in stealth, but just for the public in some way. “I’m doing a stealth company,” and people are like, “What is it?” and whatever. All the theater that happens around metrics and how they position those metrics and vanity metrics. We’ve had a huge movement over the last decade, clearly on this, where founders have become creators, like founder-as-creator culture, where they are putting out their days, and they’re looking cool while building their startup. In some ways, this has created a whole fad around entrepreneurship, where it feels cool all the time, where, to be honest, it’s very painful.

Bertrand Schmitt

Yes, it’s not so easy.

Nuno Gonçalves Pedro

It’s a very painful activity being an entrepreneur. It should be. It’s cool, but it’s cool only after the fact, so to speak, or after you get to scale. There’s a lot happening in that field where people are just, “Oh, I am reaching this kind of metrics.” as we’ve discussed as well, not all of those metrics are actually truthful, or they’re lying in some ways, or they’re obfuscating the fact that some of these metrics don’t quite mean what they say they mean.

ARR has become the most bastardized, probably, of all metrics. It used to be annualized contracted revenue, so revenue that has been already contracted. Now it’s like, “Oh, I’m just picking up my monthly revenue, and I’m multiplying it times 12 and that’s my ARR.” We’re like, “Well, what about churn? What about—are the contracts’ month to month or not?” now we as investors have to ask a lot of questions that we didn’t have to ask before because people are like using this as, “Oh, it’s a beautiful metric. Let me tell you where I am on ARR.” But at the same time, it feels like the metric is no longer really a truthful metric in some ways.

Bertrand Schmitt

Yeah, but I feel at the same time investors are more and more used about the bullshit in some ways. They know which questions to ask what to ask. If you bullshit too much your way, I’m not sure it’s good in front of investors, to be frank, because at some point you’re expecting people who are a bit mature, who know what they’re talking about. If they don’t give you proper ARR and stuff I can see that. The more questions, discussions you have about the quality of ARR, the reality of ARR, the less interest there is.

Honestly, it’s easy to get momentum with some metrics, but it’s also very easy to lose the momentum when investors discover that these metrics are not real. Or that the entrepreneur doesn’t even understand what’s going on behind these metrics. I don’t know. I agree that there has been a rise in bullshit metrics, but I think there has been also a corresponding rise in attention to details from investors to figure that out and ask the right questions. In some ways, we might all be losing time these days, but I think everyone more or less knows what to ask.

Nuno Gonçalves Pedro

Where would all that productivity we’re getting from AI go to? That’s great. Now we have all this productivity, we can waste more time. But being less facetious, there’s some studies. There was a Purdue study that was using Crunchbase and LinkedIn data that found that the average number of founders’ LinkedIn followers—so the LinkedIn followers that founders have on LinkedIn—was the strongest predictor of funds raised. Then there was some earlier Wharton, HBS, and Indiana study that found that startups active on social media were more likely to close funding rounds and raise more capital. There’s other stuff out there that’s showing similar things, like: if you post more, you raise more.

Now, me being old school, I would point people to From Good to Great by Jim Collins, that has a leadership chapter that shows the opposite: that the level five guys were less than half on media than the level fours. They were clearly—because they were level five—they were outperforming dramatically level four leaders, level five obviously being the best. I’m not sure. It is also interesting to look at what’s happening with the funnel. Less and less firms are able to raise Series A.

Maybe that kind of hotness is useful if you’re very early onto your journey. Very early numbers, in some cases, even if you haven’t launched yet. But once again, the rubber meets the road, and you’ve raised a little bit of institutional money—a pre-seed, maybe even a seed—at some point, you’re going to be judged. That seed-to-A crash line, so to speak. Your ability to jump that chasm is pretty, pretty vital. The reality is actually the number of seed-funded companies that are raising institutional Series A within 24 months has gone down dramatically. 15-20%, down from 25-30% back in 2018 cohorts. I think actually the numbers—the real numbers—are even worse than that. These are numbers that are probably not fully randomized.

Bertrand Schmitt

Yeah, I would just say that I feel, in a way, we live in an age of AI where quite a lot of moats disappear. The best technology teams, science teams, the best developers. Your 10x engineer is now an AI. It’s not an engineer. You can argue that a 10x engineer with AI is still the best, and that’s probably true. My point is that I think that we are moving to a space where a lot of moats are disappearing, and having a really good communication, visibility is actually really a moat. Between a few companies that are going to fight for the same type of clients, financing, or the like, being better at communication can really make a difference.

I think for me the question would be, can you do that while at the same time building a solid company and delivering real results? If you can combine both, I think that’s really huge. If it’s all about the communication, but not the results, we have a huge problem. Usually these persons might have raised a lot of money, so it can be trouble. But I think I would be careful.

I would say these days, from my perspective, communication—proprietary pipeline of investors, of prospects, customers, because you are very visible on X, on LinkedIn—I think it has some real value in order to move much faster than others in your space. I would still be careful. I think there is real value there. I think the question is not to overdo it. If you promise way too much than what you can deliver, it’s going to be painful because at some point you have to deliver the goods, both to clients and to investors.

Nuno Gonçalves Pedro

Everyone’s becoming more litigious, right? Venture capital firms, even angels, private equity firms, etc. If they feel there was something that was positioned to them that is absolutely fake, or even touches fraud—they will not hesitate to take these companies to court. Part of it, I’m not sure, is that we have a lot more fraudulent activity going on or not. Maybe we do. Part of it is I feel investors now are more aggressive. They’re like, “I don’t give a shit. You misrepresented the company. You gave me numbers that are not correct, from exaggerating the number of users you had. Many of them were bots.”

That’s happened at least a couple of times that I can recall recently. Some of them are still active litigations. We remember Builder.ai, which initially was positioned as—when it came out, the first story—it was initially positioned that there were a bunch of people actually pretending to be AI. That seems to have been debunked. I’m not sure fully or partially. Maybe that’s no longer the issue. Maybe that was just a hatchet job. Then there were some discussions that they actually inflated their sales and/or revenues. All of that’s going on.

Again, I’m not sure that it’s true that we have worse and worse companies in terms of how they represent their numbers and in terms of how they go to market. But there was a huge incentive, and I feel that’s the point of the episode that we’re doing today.

There’s a huge incentive for people to, again, create smoke and mirrors, to show brilliant vanity metrics, to show growth, to show a beauty of a message that sometimes can’t be really replicated in real life. Which means there has to be necessarily a delta, the reality distortion field that you, Bertrand, were alluding to earlier. If you go too far with it and if you go into communication with it, you’re going to get kicked in the butt. Sorry, but people are going to say, “Hey, you lied.”

Bertrand Schmitt

I don’t know you, but I think things started to change after the fraud around Theranos. If you remember Elizabeth Holmes, she was the typical entrepreneur thinking that, you know, you can fake it until you make it. Maybe there is a bit of truth to that, but to that level of true misrepresentation, fraud, and threatening people— =this was crazy. What she did, what she built—it was an empire of fraud, basically.

I think this was probably for me a turning point. I think at some point people realized, “Hey, taking risks is one thing. Maybe a bit of faking it until you make it, but at some point there is a yellow line that you should not cross.” She went really big, if you remember her presentation. Her style was copied from Steve Jobs. She was a great communicant, but behind the scenes. The product was not working. The product was BS.

I think it’s a good reminder. You don’t want your marketing to be too far ahead of your product. That’s something I always tell entrepreneurs to be careful. You want your marketing to be rooted in some level of reality. If it’s totally disconnected, there will be trouble with your team, with your customers, with your investors.

Nuno Gonçalves Pedro

The Megaphone

Moving to the megaphone. Now we’re at a different scale. Now we’re talking about VC firms. Obviously, we’ve talked about A16Z and their communication machine in the past, but there are others. Harry Stebbings with 20VC, right? He built a podcast, and then he built a firm around the podcast, which is great, and Harry has done incredibly well around it.

But talk about communication leading to substantive fund management, VC firms obviously have a lot of interest in talking positively about their investments. They can do that in many ways. They can do that by speaking about their companies in high regard—their portfolio companies in high regard. They can package some of their thought leadership and their thinking and their market scans and everything that they publish out there, or even some of their sound bites—in some cases even policy. Again, A16Z. In a way, that sort of illustrates a world in which their companies, their startups, their portfolio companies are the necessary solution.

At that point in time, we’re entering a field of communications that’s a little bit hazy, right? Bertrand, it’s a little bit like you’re saying, “Okay, this is the megatrend, this is what’s happening.” Magically, this fits very well with the companies you invested in. All of them fit well into this.

Bertrand Schmitt

Yes, it makes sense, I would say, as long as you disclose a minimum. But I would say another thing, when you think about it, that megaphone era—the VC content machines—in some ways, it’s also connected to the death of mainstream media and the death of the startup media ecosystem. Think about it. In the past, you had TechCrunch; you had all of this ecosystem of startup-focused media.

The Information is still alive and kicking, I would say. But at some point, there was a realization that this media ecosystem for startups was not so friendly anymore. For some reason, they decided to move from celebrating startups, talking about startups, to being very angry and nasty people. At some point, I think entrepreneurs and VCs realized they don’t want to depend on an untrusted third party that seems to hate them in some ways.

It was not the case between 2010 to 2014, 2015, I would say. But definitely, the switch was done in 2020. The marriage was over between the startup media and the rest of the ecosystem. I think at some point, there was a realization that you have to go direct. You cannot let either the mainstream media or the startup media be in between you as an entrepreneur or VC and investors and LPs or the government or whatever. At some point, there was this realization that, “Okay, we need to go direct,” so my point is that I think the two are connected. If the startup media ecosystem had done its job, we might not be there anymore. Maybe it was going to happen anyway at some point, I don’t know. But there is a clear causation here.

Nuno Gonçalves Pedro

Yes, and A16Z is well known. We’ve discussed it also in previous episodes. They went very far at the beginning. They figured out that the best option for people to come to them was to really create a lot of buzz around themselves, which was incredibly well done. Clearly one of the pioneers in the space.

Margit Wennmachers was a well-known PR person in the Bay Area with her firm Outcast, early on helping them. At some point, they just decided to try to get her in-house, which they did. To do that, they actually had to buy her firm, so there are people today who say that A16Z has more than 70 people on the marketing team. I wouldn’t be shocked if that’s the case. I don’t know the exact number, but I wouldn’t be shocked if it’s more than 70—7-0.

Bertrand Schmitt

That’s a big number.

Nuno Gonçalves Pedro

I mean, when most firms would have one to two—I mean, just to be clear, most VC firms in Silicon Valley still, and by most, I mean easily more than 90%, numbers from 2018 would have told us 95% of all VC firms in the Bay Area have less than 10 people overall. These guys have 70 people on marketing. You know, if they do, it’s incredible. Incredible. I think that playbook worked really well for them. It has worked well for many of their startups. At some point, there’s dilution of messaging, I would feel. But certainly, there’s an element of propelling the message, of thought leadership.

We’re not trying to say A16Z has done something wrong. They were pioneers. They did something very well. They changed the market. The same way that Harry at 20VC has done incredibly well and built his fund management business out of his podcast. Incredibly smart and really propelling him to be central in many cases in early stage. But then there’s a question on, does this work? It works for VCs for sure. Because a lot of the capital committed is going to very few VCs right now, A16Z being one of the brilliant examples of that. The one that has propelled the most in terms of assets under management, coming out of a 2008 or 2009 fund, if I recall correctly. They’ve done incredibly well, so the key seems to be the word around marketing, PR, and how you present yourself as a VC firm, but also on how you introduce your portfolio companies.

For those listening to us, if you are listening to podcasts or reading articles from any VC firm, from any venture capitalist—including myself, just to be honest, or Bertrand, probably—just take into account that at the end of the day, we are not fundamentally neutral. It’s not that we’re trying to lie, or we’re trying to prop up our assets. It’s maybe the other way around—is that there was a system of beliefs that led us to invest in specific companies, and that system of beliefs has manifested itself in those portfolio companies. Obviously, we stand by those systems of beliefs going forward. We’re not necessarily neutral parties. We have beliefs. We have things that we believe in terms of how the future is going to pan out, architecturally certain markets are going to pan out, et cetera.

Bertrand Schmitt

I think you’re right to be honest and transparent here, you know, in terms of a system of beliefs. But at the end of the day, it’s true of every company. News media, startup media, financial media—they have their own system of beliefs, their own approach. They have their own goals. For an individual journalist, it might be to get a special award or something.

They might not care if some startups have been pushed into the mud, if there was a mistake, because at the end of the day, their objective is not at all about the startups. It’s about their own personal or company interest, so I think we have to recognize that every actor has some incentives, has their own logic, and you have to be aware and careful about that.

Nuno Gonçalves Pedro

Communications—the same as PR, the same as news—is actually not necessarily factually correct. We should know that by now, because we all listen to the news, we all read news. At some point, we may have been involved in pieces of news, and we come to the realization that there’s no such thing as full neutrality.

Again, the ability to triangulate different data sources, different news, different articles, different perspectives is very valuable. The ability to form one’s own opinion is very valuable. Again, we’re not saying that communication and communication machines are not important. They are. The world has been built on advertising, PR, and many other things. But you should look at it with a grain of salt, and particularly if you’re trying to avoid stupid decisions like investing in a company without having a clear view on what the company is capable of, et cetera.

Maybe moving to our last section: the big guys—the guys that, when they talk, we have to listen, and the prices of their stock change everything for us. They change our 401(k)s, they change our pensions, they change whatever. Which are the big tech companies in particular. Talk about a bunch of people that are, I wouldn’t say not to be trusted, but not to be trusted.

Bertrand Schmitt

Yes, and I started about Apple and Steve Jobs and his famous reality distortion field. You talk rightly so about Microsoft. At some point, they were the king of vaporware. Things have changed, obviously, for these companies. We’re talking about 20 years ago or something. But obviously, these guys are masters at it. I would say what has changed for me is the advent of AI.

The advent of AI has changed quite a bit what’s happening. Instead of being usually relatively positive—I would say not really focused on trying to get regulation; on the contrary, trying to get regulation away—we have seen, I would say, a disturbing pattern, especially from Anthropic, but also from others like OpenAI, trying to scare, using scare tactics.

This is quite new. I don’t remember scare tactics being part of the arsenal in Silicon Valley before. That, for me, is quite shocking and surprising. At the same time, we see actually pushback from the bigger guys, from the Nvidia of the world, Microsoft of the world, and trying to present their part of the story and also trying to push back on some regulations. I think it’s a big difference. It’s quite surprising.

I would say, sadly, I believe that some of the AI players trying to scare, trying to regulate, in some ways, are finding that the traditional media is a great ally in that direction. Because as everyone knows, the traditional media thrive on fear, on sharing scary news. They are probably quite pro-regulation as well. In a way, it has worked very well. “I give you content to scare your audience,” and this has been some sort of vicious cycle, I would say. I don’t know your take, you know, but I’ve been quite saddened by what happened in the past 12 months, 18 months, especially regarding AI.

Nuno Gonçalves Pedro

Yes, and it’s a little bit related to the episode we did on the ruling of the doors. All of this has gone geopolitical—how regulation’s done de facto, certainly in the US, and how there’s this knee-jerk reaction to a lot of things happening in the market at the same time, and honestly, with incestuous relationships between different areas, where political power has incestuous relationships with tech right now, it’s very clear that there’s a lot of stuff happening between DC and Silicon Valley that didn’t happen before quite in the same way, and so when that happens, people feel that they can do whatever they want. Like, “Oh, let’s create fear, uncertainty, and doubt—FUD for short—because that’s a good way of moving not just consumers, not just the price of our stock,” and this is mostly private stocks right now.

Obviously, there are a couple of other guys that are in the public stock market, like SpaceX. “But let’s not just move people around our stock and around how they perceive our company, but also let’s move governmental decisions. Let’s move industrial policy, let’s move regulation.” I think that’s how the game has changed, and that’s the fear, uncertainty, and doubt taken in a new direction. There’s been FUD. Again, Microsoft probably was the granddaddies of that, but there’s been FUD for a long time. But now FUD is at a different dimension. It’s geopolitical. It makes people that have no clue of what Anthropic and OpenAI actually do—it makes them scared about AI, et cetera.

Bertrand Schmitt

You know, the difference is that even though I agree with you—Microsoft, I mean, this acronym of FUD was early on associated with Microsoft, in particular, maybe 20, 25 years ago, and FUD stood for fear, uncertainty, and doubt. What’s crazy to me is that, in the past, Microsoft would use that against their competitors, against their smaller competitors. It was not about themselves or their own industry.

That’s not what the fear was—the scare was targeted to. Where here, you have Anthropic saying, “Hey, regulators, we are so dangerous—we are going to destroy the world as we know it,” so it’s completely different. It’s not like, “I’m going to point to the other guy saying he’s doing bad stuff.” No, I’m pointing at myself, and I’m saying, “If you don’t regulate me, I’m going to destroy the world.”

For me, it’s a pretty shocking development because with traditional FUD, you could understand the business logic. Here, it’s completely different. It’s like, “Are these guys a cult? Are they crazy? What’s truly going on behind the scenes?” Obviously, it looks like they want regulation because they believe they will benefit from regulation. But it’s just surprising for me. It’s a very different approach from what we saw decades ago.

Nuno Gonçalves Pedro

My best guess is it’s twofold. One, it’s a well-known game in the US. You guys can check out Gurley—Bill Gurley, former general partner at Benchmark, has a couple of videos on YouTube about this: regulatory capture. That’s the first intent. They want to do regulatory capture because they know if there’s regulatory capture, they basically get within that framework, and then everyone else is outside. Therefore, the people that are outside won’t have a chance to compete with them, so in some ways, they become the oligopoly’s forces in those markets. They’re not doing this for the benefit of others. They’re doing this against others for them to really be framed and protected. That’s one play that they could be doing. I assume some of them are.

The second play that they can be doing is obviously the notion that preemptive regulation eliminates the regulatory risk behind the company, and it allows them to go public without that risk. Put in a different way, if I want to IPO as a company, and I’m trying to pitch myself to underwriters and then to the market generally, once the company is actually fully public so that retail investors can invest in me, I want to eliminate as many risks as possible, and one of the core risks is still regulation around AI.

If AI gets regulated in a way that’s favorable to me, I eliminate that risk. It’s well understood, or at least I distill it to something that’s very understandable, and therefore that gets priced into my stock with a premium, so my stock goes up, so to speak. I think those are the two intentions, unless I’m missing something a little bit more nefarious. Those are the two intentions that we have today in the market.

Bertrand Schmitt

In the past, the biggest company used to try to push back regulations as much as they could, and only when they truly became mainstream, and decades had passed—only at some point—they might start to push for regulations in some ways. To see a company so young looking to get regulated, and not just regulated, but, you know, scaring everyone about what they could do or achieve one day—it’s scary. Think about the tobacco industry, for instance. They were not saying, “You know what, people are going to die by smoking tobacco. We need to be regulated.” No, they tried to hide it. They tried to fight it. At some point, they had no choice but to accept it, but it took them decades to get there.

Nuno Gonçalves Pedro

But maybe that’s the accepted orthodoxy today, or doctrine today. If you get regulated, and you can control or be heavily involved in the process of regulation, you reap a lot more benefits than avoiding regulation in the first place. Maybe back to your Philip Morris and whatever big tobacco example—if they had self-regulated and pushed for actual, heavily, more heavily regulated stuff earlier on, maybe they would have ended up better.

That, I think, is what’s happening in tech. In tech, I think there’s the notion right now—if you’d ask 9 out of 10 general counsels, chief regulatory officers out there in tech, 9 out of 10 would probably say, “Actually, if we have scale, we’d prefer to be regulated and be involved in that process,” and shape it.

Bertrand Schmitt

You might be right. I just feel pretty sad if an industry goes there. It’s pretty scary. It’s not good. It’s not the way capitalism should work. It’s a capitalism of convenience with the government. Personally, I believe we should have a small government, not a big one involving every decision of your life, either as an individual or a business. I think, ultimately, it’s bad. It goes contrary to the ideals of a country like the US and many other countries.

Yes, maybe that’s what’s happening. I just don’t think it’s right. I don’t think it’s good. Maybe on another topic, beyond AI, we cannot not talk about Elon Musk. He has been one of the most talented communicators in history. You might say he’s not a communicator, but he is. Obviously, he’s everywhere on X. He acquired Twitter as well. But he’s pretty impressive. I remember years ago—maybe 10 years ago—when he was pushing that full autonomy is right around the corner for self-driving, that we will have a million robotaxis in 2020.

Just the latest launch of the new Roadster was pretty interesting. Personally, I love the launch, but it’s very interesting to notice that, actually, they are not going to sell that rocket version of the car. It’s invite-only. You can buy the normal version, you can buy the booster version, but the true rocket version? Not clear who will be able to buy it. Obviously, there’s a question that it has been in development for 10 years. It might not be released until maybe 2 years, maybe more. It’s very impressive how he’s able to manage, but at the same time, capture the attention, get people interested.

Clearly, it’s a part for him of doing business. Why does he have cars with strange shapes, strange features? I think a lot of it has to go with communication. He could have done a proper traditional SUV. He doesn’t have one in his lineup. It will sell like hotcakes. But I guess his mindset is, “I don’t want to sell something that looks like somebody else’s car,” so he has to be very special, very marketing-oriented, in order not to spend much on marketing at the end of the day.

Nuno Gonçalves Pedro

He can’t afford to spend as much money on marketing as he wants right now, but he is not. Talk about a one-man marketing machine, right? This is the most incredible.

Bertrand Schmitt

Yes, he’s a one-man marketing machine. He’s cheap. He’s cheap.

Nuno Gonçalves Pedro

A one-man marketing machine that moves the price of all his stocks, moves the perspective on him. A group of fanboys and fan girls around him is silly. Everyone loves Elon.

Bertrand Schmitt

Not everyone, but yeah.

Nuno Gonçalves Pedro

Some hate him, but most people love him and idolize him. Talk about the creation of a cult. Clearly, he could create a cult tomorrow. Maybe he already has. It’s incredible what he’s done. To a certain extent, we have to watch in awe. He is incredibly capable of running his companies and creating new products and being ahead of the game. He can lie and say, “Oh, the cars will be here in 2 years,” and it takes 10 years, or “Robotaxis are going to be here,” and they’re not. Full self-driving, unsupervised is here by whatever, 2025, or whatever was the number that he gave, and now they’re talking about Q4 this year. It’s like nobody believes anymore. Robotaxis were 2020. Initially, full autonomy was 2018. It is what it is. He is that kind of one-man marketing machine. Not many around today. Jobs was that kind of person, but he was more really focused on products. He wasn’t tweeting. He wasn’t out there manifesting image. It was all around demos. Even that—the demo space has now changed a lot with BS demos. Gemini launched a video, which apparently was assembled using still image frames back in 2023, and then a prompting via text. It was just totally staged.

Apple, from what I recall, no longer does live tech demos. Those were the days. Actually, I would allege the first demo of the iPhone by Steve Jobs wasn’t fully live. There were people in the background doing stuff—live-ish. It’s the games people play, even demos. They’re like, “Well, that’s not really… Is that true?” Probably not. Cool. Beautiful video or beautiful thing. They’ve been able to put all of that stuff in, but all of that is no longer true. Then the last, but not the least, thing that comes to my mind is—talk about when a company actually rebrands itself because a clear bet, financial and otherwise, but really talk about a communication bet that leads to the rebranding of a company. Meta being a perfect example of that—going from Facebook to Meta because of their bet on Metaverse, and then their unit showing losses over and over again to what seems to be 80-something billion in cumulative operating losses against a little bit over 10 billion in total revenue since late 2020. Now, effectively, another rebranding going on where Meta now is an AI company. That’s what Zuck has been telling us.

That’s the ultimate branding. It’s when you change your company name to align with what you’re focused on. They did definitely put money behind it. It didn’t quite work out probably the way that he would have expected it.

Bertrand Schmitt

You can notice SpaceX actually changed its name—from SpaceX to SpaceX AI to SpaceX xAI, potentially a merger with Tesla. Some company names are changing. It made me think about this trucking company—electric truck companies. If you remember Nikola, at some point the CEO was accused of fraud. I think it was an example of someone who really promised way more than was real and said stuff that were clearly untrue. I think, at some point, again, if you have too big a gap between communication and reality, it’s coming back at you at some point. But Elon Musk, at some point, has managed to always deliver. Yes, it might be 2 years late, 5 years late, 10 years late, but in the meantime, he delivers value to customers. He sells his stuff. Some might be frustrated, refuse to listen to his next prognostics, but at the same time, he managed to always bridge the gap, which is pretty rare.

Nuno Gonçalves Pedro

Now communication has gone mainstream—like, AI is mentioned all across the board. Recently, for Q1 2026 earnings calls, FactSet basically found that AI was cited in 337 of S&P 500 earnings calls. 337 out of the S&P 500—that’s incredible. AI has to be there, otherwise you get killed. Your stock goes down. If you’re not doing anything around AI, you’re going to go down.

Bertrand Schmitt

It’s quite clear. The next big thing.

Nuno Gonçalves Pedro

Honestly, the last play is all the circular economy stuff that we’ve been mentioning in previous episodes, where someone announces, “I’m investing in your company 30 billion,” and then you realize it’s left pocket, right pocket—like Nvidia announcing it would invest 100 billion in OpenAI. It later scaled it down, I think, to 30, and saying, “What is that money going for?” Well, magically, it’s going to go towards OpenAI buying Nvidia chips. It’s like, cool, it’s like some sort of financing agreement. There’s worse out there. There’s now third-, fourth-party circular agreements—not just two-party circular agreements, but like three-, four-party circular agreements.

Bertrand Schmitt

You could argue it makes sense that Nvidia in some ways had to kickstart some of this, because if you are a big AI lab, your biggest line of spend is GPUs. Being there to finance GPUs at a time when you are making lots of profits because you can increase your price makes sense. Why just sell when you can invest and sell? I think that’s a smart move. History has proven that the investment and support for Anthropic, for OpenAI, were very good business investments. At some point, you could argue it’s a recognition that even them cannot do everything. They realize that having that focus on the GPUs is key. While Nvidia is definitely building more and more open source, open-weight models, they know that the frontier is with Frontier Labs. Therefore, you can as well invest in them and not just sell them the GPUs.

Nuno Gonçalves Pedro

Regulators are paying attention. I think they should have been paying attention for a while—SEC being obviously one of the obvious ones to be paying attention. They’ve issued apparently 92 AI-related comment letters to 56 companies since 2021. They’ve mentioned that anything regarding regulation around statements of AI capabilities is a priority. I think this is a catch-up game. It’s very difficult for any kind of regulator—SEC or anyone else—to really keep track of everything. Even if they know all the announcements, then how would they know if the capabilities are true or not? We’ve seen it in the past. Elon doing stuff that clearly moved his stock price, nothing happened—get a slap on the wrist and let’s move on. I don’t know. I feel it’s interesting. There should be more active views, in particular, on anything that links to retail investors and consumers. It should be a little bit more than a slap on the wrist if people are outright lying.

Bertrand Schmitt

I think it’s a mix. I think at some point, if you remember in the past, you couldn’t even tweet as a tech CEO. We saw the SEC coming at you if you dare to make a mistake. I think it was important to, step-by-step, make it easier to communicate on Twitter, on Facebook. They were not considered official channels you could use to communicate to your investors, when actually they were the most efficient channels to communicate to your investors. I think, definitely, they used to be quite late. At the same time, you also want to root out true fraud. Again, we take the case of Nikola. You want to make sure that there is some backing of your capability. At some point, you should explain more what is it that you sell. But that’s true that at some point, when you see something called full self-driving, and you realize that it doesn’t do full self-driving—you have to put your hands on the wheel and this and that—it’s surprising that it managed to keep that way for a while. At the same time, we might be finally there in terms of full self-driving.

Nuno Gonçalves Pedro

In some ways, we’ve gone full circle in the episode because we’ve had, “Why do startups fake it till they make it?” Well, because big companies also fake it till they make it on products and new areas of focus, et cetera. VCs as well. The ultimate manifestation of that, guess what it is? Communication—how you communicate with the market, how you present yourself in social media, how do you do PR, anything. Even now, founders’ tweets and whatnot.

Bertrand Schmitt

Our Take & Conclusion

To conclude this episode, it’s clear that in some ways, there is nothing new under the sun. Communication is key. At the same time, communication has evolved quite a lot. It’s become more and more direct, more and more personal. Less and less middlemen are in this game. It’s more and more the big CEO talking directly to his investors, to his customers, directly through social media. You are in this situation where if the big guys are over-communicating, over-promising, then for sure, as a startup, can’t you do the same? Shouldn’t you do the same if you want to stand a chance and to be able to keep the fight? That’s where we are going. At the same time, there is a fine line. If you go to fraud, they will come at you, rightfully so because there is only so much you can promise to employees, investors, entrepreneurs, customers. You need to make sure this doesn’t go too far so that we keep a strong, vibrant ecosystem. Thank you, Nuno.

Nuno Gonçalves Pedro

Thank you, Bertrand.

Weitere Episoden von „Tech Deciphered“